Observing Evolution Over 40,000 Generations 461
Last year we discussed the work of Richard Lenski, who has been breeding E. coli for 21 years in a laboratory in Michigan. Then, the news was that Lenski's lab had caught direct, reproducible evidence of a genetic mutation with functional consequences for an organism. Now Lenski's lab has published in Nature a major study comparing adaptive and random genetic changes in 40,000 generations of E. coli (abstract here). "Early changes in the bacteria appeared to be largely adaptive, helping them be more successful in their environment. 'The genome was evolving along at a surprisingly constant rate, even as the adaptation of the bacteria slowed down,' [Lenski] noted. 'But then suddenly the mutation rate jumped way up, and a new dynamic relationship was established.' By generation 20,000, for example, the group found that some 45 genetic mutations had occurred, but 6,000 generations later a genetic mutation in the metabolism arose and sparked a rapid increase in the number of mutations so that by generation 40,000, some 653 mutations had occurred. Unlike the earlier changes, many of these later mutations appeared to be more random and neutral. The long-awaited findings show that calculating rates and types of evolutionary change may be even more difficult to do without a rich data set."
Re:goodbye creationists (Score:3, Informative)
Not yet.
The creationists will blindly and steadfastly cling to their mysticism-based pseudoscience until two chimps mate and produce a homo sapiens offspring.
Which of course is not how evolution works.
Re:Creationists response: (Score:5, Informative)
51% of people believe god created man as he is.
30% said god created us and we can evolve
15% say humans evolved with out god.
These figures are a terrifying example of humans ability to deny what should be blatantly obvious. If we can do this imagine how many things people must get completely wrong no matter the level of obviousness.
Re:Someone needs to tag this "Inbreeding".. (Score:2, Informative)
Is that are joke or are you intentionally dense? http://en.wikipedia.org/wiki/Inbreeding [wikipedia.org]
"Inbreeding has a variety of consequences. Allele exposure can cause genes to be expressed that are not otherwise expressed. This fact, combined with the fact that most mutations are recessive may indicate that inbreeding drives evolution. Speciation, a key process in evolution, depends on reproductive barriers, a necessary feature of which is inbreeding."
The mutation process here is driven by inbreeding and keeping the population isolated.
Re:uhh? (Score:5, Informative)
Yes. You are missing the fact that this experiment has been running for the last 20+ years. Time is the major factor here. Furthermore, they did a bit more than simply comparing the DNA from the current strain to the original strain. THey kept samples of strains of the bacteria every 500 generations or so and compared them. Even running parallel experiments using these stored strains allowing them to effectively repeat the experiment in order to understand the evolution of the new metabolic pathway allowing for the utilisation of Citrate.
Re:uhh? (Score:5, Informative)
A main purpose of the study is to investigate evolution of phenotypes, not just genomes--- i.e. how the functions and capabilities of bacteria change over generations due to evolution. Just showing there was a change in the genetic sequence doesn't do that, since it might be a change that isn't expressed.
Re:uhh? (Score:4, Informative)
That would be one way to go about it, all right. There are a couple of problems though. Current state of the art DNA sequencing runs somewhere in the range of a few tens of thousands per (for humans, perhaps a bit less for something like E Coli). That's a technological advancement, all right, considering when they first started sequencing genomes it was a billion dollar project. It's also not instantaneous. Much faster than it used to be (years or decades) but not instant. Note that the samples he's looking at are ones that have been frozen periodically over the last twenty years. Apparently the price of sequencing genomes has dropped to the point where his lab has the funding to actually do it now.
The diff part isn't trivial either. The genome for E. Coli is around 5 million base pairs long, which doesn't sound like much, if you're just looking for point mutations. The problem is, there are lots of other things that can happen to a genome besides point mutations. Genes can hop around or get copied into the wrong location, which you might count as no mutation, or one mutation, but either way you still have to figure out where it came from. Also, although E. Coli reproduce asexually, they do share genetic information through conjugation, so you get gene shuffling that way. There's also at least some genetic diversity in the colony, meaning you'll be dealing with several different genomes.
Once you've worked all that out, it's not all that interesting just to look at now vs. then. If you wanted to do that you could go dig frozen bacteria out of ice cores or something. The point of this experiment was to be able to watch as the genome changed. So you have to do lots and lots of comparisons, from samples taken at different times (every 500 generations, IIRC, meaning about 80 timepoints). Oh, and there were multiple, isolated populations.
On top of all that, what's really interesting is functional changes. Counting mutations is fine and all, but you really want to know what (if anything) those mutations are doing. The headline event was a mutation that allowed the E Coli to metabolize citrate, for example.
Re:hmmm (Score:5, Informative)
ALL mutations are random. If they are advantageous, great, than it is likely that they will be passed along.
Re:Creationists response: (Score:5, Informative)
http://rationalwiki.com/wiki/Lenski_affair [rationalwiki.com]
Re:hmmm (Score:4, Informative)
Evolution is a two-step process - the first part is the production of mutations, which is a random process (and, given how finely balanced organisms are, the majority of these random events will probably be negative, on balance). The second part is selection - if there is genuine competition between these strains, then the beneficial mutations will be selected, so the fact that they are relatively rare will have little effect on their eventual domination of the population.
Re:Yes, that Lenski (Score:4, Informative)
Re:hmmm (Score:1, Informative)
Re:Someone needs to tag this "Inbreeding".. (Score:4, Informative)
Is that are joke or are you intentionally dense? http://en.wikipedia.org/wiki/Inbreeding [wikipedia.org].
You should be reading this page instead http://en.wikipedia.org/wiki/Asexual_reproduction [wikipedia.org]
Re:hmmm (Score:3, Informative)
You missed the evolution of the ability to metabolize citric acid.
There's no luck involved here- there's just a mindbogglingly high number of mutations and tests of those mutations over the course of history. You say "even over the lifetime of the universe", but it's unclear to me that you really understand just how large a number of generations there have been even since the rise of life on Earth, let alone the entire span of the universe. That's a lot of individuals, a lot of generations, a *lot* of mutations, and therefore an enormous well of opportunities for change.
Re:hmmm (Score:2, Informative)
No such thing? Are you retarded or willfully ignorant: http://www.talkorigins.org/faqs/macroevolution.html
The only difference between macro- and micro- evolution is the time-line.
Each little step/mutation is EVOLUTION, to split the changes into "micro" and "macro" is to diminish the meaning.
You could have dogs evolve into whales, but if you looked at each mutation individually, you could dismiss it as "micro-evolution".
Re:Evolutive pressure? (Score:2, Informative)
There is a step from "DNA mutation" to "Evolution", and that is adaptation to the medium. Did the mutations change the fenotype (the external aspect/behaviour) to something more adapted? Were set certains goals (for example, putting them in a medium less than ideal for the original strain, but to which its survivors have adapted) or the surviving changes did not affect at all at the species?
I think what you're trying to ask is: "Was the selective pressure determined to be in response to stimuli versus a random occurance?"
The authors cover the difference between neutral drift and selective mutations which increase fitness throughout the paper.
Specifically in answer to your question, though, is the following from the expanded methods & materials:
"We performed Luria–Delbrück fluctuation tests33 to confirm that the Ara-1 population evolved an elevated mutation rate. Bacteria were revived from frozen stocks by growth overnight in LB medium. After dilution and 24 h of re-growth in Davis minimal medium supplemented with 25 mg l-1 glucose, we inoculated 24 replicate 10-ml cultures of Davis minimal medium with 250 mg l-1 glucose with 100–1,000 cells. After 24 h of growth to stationary phase, these cultures were concentrated by centrifugation and plated on LB agar containing 20 mug ml-1 nalidixic acid."
Re:Yes, that Lenski (Score:3, Informative)
http://en.wikipedia.org/wiki/Conservapedia [wikipedia.org]
Re:hmmm (Score:5, Informative)
"Look, I believe in evolution, but never has there been found a parent species to something alive today. In other words, scientists can not point at any two distinct species, living or extinct, plant or animal, and say that this species evolved directly from that one."
Of course not. That's kind of like pointing to two leaves on a tree and saying one leaf came from the other. It doesn't work that way. They are both on the terminations of the branches, and the node where they branched into two is in the past. Ordinarily, the common ancestor is long dead. The nice thing with these E. coli is that the researchers kept a portion of the ancestral population intact, and the specimens are clones, so while not the actual ancestor of the lineage that kept going, they are genetically identical.
There are plenty of fossils that are close to branch points, and as more fossils are found there are still plenty of gaps left, as there always will be, but the changes necessary to span those gaps get smaller and smaller as the sampling improves. For example, Anchiornis [wikipedia.org] was just discovered in the last couple of years, and a new specimen described a few weeks ago. Dinosaur? Bird? It's rather arbitrary to decide. It's either a wing-clawed, long-tailed, toothed bird like no modern bird, or it is a flight-feathered, gliding dinosaur. As if they were the leaves on a tree, birds and reptiles look distinct now, but follow the branches back far enough and they get mighty blurred together. This is hardly an isolated example.
There are fish that look so tetrapod-like that when the skull was initially found separately they thought it was a tetrapod [devoniantimes.org]. Then workers found the rest of the body and realized it was a fish. There are other tetrapod-like fish, such as Tiktaalik [uchicago.edu]. But go back 100 years and these species weren't known at all.
I really don't know what more skeptics are expecting. Perfection? It won't happen. It's not like we'll ever have every twig on the tree. Good fossils are rare. But the statistical pattern with increased sampling is quite robust.
Re:hmmm (Score:2, Informative)
You know, it only took me 10 seconds to google for an example, you could have tried a bit harder. "speciation observed in salamanders":
http://www.santarosa.edu/lifesciences2/ensatina2.htm
A species observed to lose reproductive compatibility between populations separated by geographic barriers.
Re:hmmm (Score:3, Informative)
They already demonstrated the E.Coli bacterium evolving the ability to metabolize citric acid... that makes it a new kid of bacterium (the inability of E.Coli to metabolize citric acid is one of its defining characteristics).
And the color white was a "defining characteristic" of swans until they found a black one.
And the black swan (Cygnus atratus) is, in fact, a separate species. So even by your own argument-by-analogy, you've agreed that the new bacteria should also be considered a new species, and thus evolution has been observed to occur.
Look, I believe in evolution, but never has there been found a parent species to something alive today.
My grandparents have all passed away, but I'm pretty sure I'm still related to my cousins.
In other words, scientists can not point at any two distinct species, living or extinct, plant or animal, and say that this species evolved directly from that one.
We've had plenty of genetic evidence from preserved material to say exactly that. But the big news about Lenski's experiment is that not only do we have living examples of a species which evolved directly from another species, and not only do we have living examples of that original species, but the scientists actually watched it happen.
Show me the fossils of the prehistoric rodent that evolved directly to today's rabbit or rat and the debate will end.
Rodent ancestors appear is the fossil record around the late Paleocene. We may not be 100% sure the actual individual fossils we have are direct ancestors of currently living rodents; they may be, say, great-great-...-great-uncle rather than great-great-...grandparent. But that doesn't matter since the existence of the latter is logically implied by the existence of the former. The debate, among anyone who actually knows what they are talking about, has been over for a very long time. The only ones claiming otherwise are the creationists.
Re:Evolution or just surving? (Score:3, Informative)
On this particular subject, and I know this is unrealistic on slashdot, it would be good to read the article. The techniques used in this study were brilliant, they are specifically designed to investigate the criticisms that are usually leveled on studies of evolution and they do so beautifully. I'll try for a quick explanation of why your criticism is invalid.
First, how the experiment worked. They put E. Coli. into dishes with a growth medium of glucose and other nutrients with glucose as the limiting nutrient (add more glucose and the maximum population will go up, take some out and the max will go down). Each and every day, for 20 years, they took 1% of this sample and put it into a new dish, each time they did so the population would boom and bust as the bacteria consumed the new glucose present in the new dish. Each day the researchers would take measurements of lots of things including cell size and total population (which would be the maximum population that the level of glucose could support) as well as occasionally freezing off a sample that could be revived later.
Now, here's where it gets interesting. Almost 20 years into he experiment, the total population (again, the max that glucose could support) suddenly shot up by a factor of 5. That's right, after nearly 40,000 generations, the maximum number of bacterium that the dish could support, suddenly increased to five times it's previous level. After looking into it, the cause was discovered to be that the E. Coli. could suddenly digest a chemical used to prepare the dishes, effectively increasing the food size by 5 times. If this ability was always present, it should have shown up decades ago and also should have shown up in one of the other 11 lines (the mutation only occurred in a single one).
Now, I said it was interesting before but I lied, this is where it gets interesting. The researchers know the average mutation rate for E. Coli. They also know approximately how many generations occurred and therefore have some idea how many mutations occurred . If the ability to digest the chemical required only one mutation, it is statistically unlikely that it would have occurred in only one of the cell lines. Obviously, this mutation was worth investigating. So the researchers thawed out old samples from that cell line and let them start evolving again. What's interesting is that samples frozen before the 20,000th generations never evolved the ability to digest the plate medium. Whereas samples taken after the 20,000th generation often did. The implication is that a mutation occurred around that time which 'primed' the cell line, so that they were then only a single mutation away from being able to digest the medium. What's also cool is that the 'priming' mutation doesn't cause any large increase in fitness. In effect, this is a perfect example of irreducible complexity evolving by natural selection (human beings creating the environment isn't artificial selection the same way breeding a dog is).
Re:hmmm (Score:5, Informative)
This also allowed the total population in that group to explode (there's now more food in total, glucose + citrate).
Another cool thing is that this smashes the "Irreducible Complexity" argument. The ability to metabolise citrate is developed by two separate mutations, which, on their own achieve nothing. Some of the populations developed the first mutation and some developed the second one, but none of them had previously developed both. This shows that the "preliminary" mutations were not harmful to the bacteria, so they just "hung around" until one of them was lucky enough to get the second mutation too.
Anyway, look up Lenski's work, I'm sure his papers (and those of his students/colleagues) are better at explaining it all than me...
Re:Creationists response: (Score:3, Informative)
Standard nonsense and moving the goal posts. 30 years ago creationists defined "macroevolution" to mean speciation. Now they use it to mean some vague broader category. Indeed, speciation at this point is so accepted that Answers in Genesis one of the largest young earth creationist ministries list the claim that speciation does not occur as an argument that creationists should not use. http://www.answersingenesis.org/get-answers/topic/arguments-we-dont-use [answersingenesis.org]. As far as I can tell, "macroevolution" when used by creationists just means "any degree of evolution that we are still stubborn enough not to accept" while "microevolution" means "any evolution which has so much overwhelming evidence that even we will accept it."
Of course, your claim about information is also wrong. The Lenski experiment given in the very subject of this is but one example. However, this gets into the non-trivial issue of how to define information. There are a variety of different mathematical definitions of information, such as Shannon information and Kolmogorov complexity. Creationists generally do not specify what form of "information" they are talking about. I'm not going to go into the details of either Shannon or Kolmogorov information theory other than to note that it is blatantly obvious under both of them that a variety of different common mutation types can increase information(for example, in Kolmogorov information theory, gene duplication will generally(although not always) increase the information level).
Instead I'm going to make a short argument that shows under any reasonable definition of information, information increase has to be possible for essentially tautological reasons. Consider a given piece of DNA with information level x. Now suppose that a mutation leads to a reduction in information to state x-k where k is some positive number. That means that the mutation back from x-k to x must add at least k information since x should have the information as x no matter how we got there. That mutation may be less likely, but is has some non-zero probability. Moreover, for most mutations that aren't extremely drastic (so say nice point mutations rather than dropping large sections of chromosomes) the mutations occur in one direction about as easily as in the other. So claiming that we don't have observed increases in information is a ridiculous claim. Any time a point mutation occurs and then we get the point mutation back in the other direction we've increased information.
Re:Mutation does not equal Evolution (Score:2, Informative)
Perhaps you should define as to what you think evolution is, before you say you don't see any.
I thought I did. E. Coli still remainds E. Coli. Perhaps I should have said I don't see anything significant about this study. I have no problem accepting that genetic mutations occur. However, it seems that this study is inferring that this is the first witnessed proof for evolution. I would be interested at the lead researchers definition of evolution.
It's a fair request that you ask. I looked it up. Good ol' Google:
http://www.talkorigins.org/faqs/evolution-definition.html [talkorigins.org]
I looked at other pages as well but it seems the most standard definition I could find was on the above page and read:
"Evolution is a process that results in heritable changes in a population spread over many generations."
Also reading further into other articles about this study it would seem that Richard Lenski and many other evolutionists hold this study as a holy grail (in comparison to other studies before it) in the debate of creation vs. evolution. If all evolution is defined as being is the heritable change in a population spread over many generations then why would there be any debate at all?
Perhaps the debate is mearly by what process did life evolve. If this study holds any significance in that debate I am not seeing one. If this study is of significance in the study of mutation then I suggest there are more clear and abundant examples elsewhere.
20 years of study for what?
I digress a bit from the original request of a definition but I believe you should now understand the point I am making.
Re:hmmm (Score:3, Informative)
Your statement that every fossil is from an evolutionary dead end is wrong however - as there are plenty of 'missing-link' fossils that are found that have a (very likely) lineage (as a species rather than individuals) to current species.
Re:hmmm (Score:1, Informative)
Re:Mutation does not equal Evolution (Score:4, Informative)
The standard method of identification of bacterial species is to determine what compounds they can eat. E. coli is defined as not being able to eat citrate. They evolved something from E. coli which can eat citrate. The new bacteria is not E. coli by the standard method of identification.
But I guess you're right... it's still just some random nigh-invisible animalcule that nobody really cares about. I mean it didn't turn into a dolphin, did it?
Re:Not Convincing to Public (Score:3, Informative)
Re:I thought this sort of thing... (Score:1, Informative)
40000 generations of humans is equal to about a million years. They just used bacteria because their financiers wanted to see quick results.
Re:hmmm (Score:1, Informative)
"No, the common ancestor is always extinct. Why?"
For fossils, yes, "almost always" or "always" depending upon definitions.
I was talking about this rather unusual experiment where a portion of the originating population is preserved and where that population happens to consist of clones. In that example the originating population that includes a common ancestor of the descendant population still exists and can be revived from the freezer. That's an exception.
However, even among fossils the issue gets fuzzy if you broaden the concept to species. If one species diverges into two, the originating population of those two species (i.e. the common ancestor) is long gone, but you might label one of the branches with the same species name as the ancient one if it appears similar. If so, is it correct to say the original species is extinct? Not exactly. You can point at it and refer to it by the same name. On the other hand, the originating common ancestor *population* sure is. The issue is muddied by the way we define "species" -- a difficult concept in itself -- and label them as they change through time. Some people would always give the two descendant species different names, but in practice that doesn't necessarily happen because the changes in one of them aren't enough to justify the new label.
"Mammals formed in the Tertiary period, about 65 million years ago."
Actually, they are much older. From the Early Jurassic [wikipedia.org], although the division between them and earlier synaspids is fairly arbitrary. In any case, you're over 100 million years off.
Ignoring the time element, I'll consider the rest of your point.
"Every mammal alive today evolved from these rodents, possibly, a single rodent. That means that every elephant alive today, evolved from a small rat-like creature 65 million years ago. We should be able to track a single line from any elephant back to a single rodent."
Yes, hypothetically, somewhere way back in ancient history there was a common ancestor. It wouldn't have been a rodent in the modern sense, but if you saw it wandering around it may have looked somewhat rodent-like, although some of the synaspids and theraspids were pretty darn big.
In short, no, we shouldn't be able to track a single line. Or at least, it should be extraordinarily difficult for two main reasons: 1) fossils of mammals are relatively rare; 2) something greater than 99% of species that ever lived are extinct, including many species that would be close but not quite on the line of interest.
What you are describing is the challenge of trying to trace an exact lineage when we are missing the great number of species that ever lived, and far more if you count individuals. It's like trying to follow a genealogy when you don't have 90% of the names. Actually, it's worse than that. You only need to look at the number of extant species (millions) and compare it to the number of fossil species (thousands) to realize how big the challenge is. At best we get a kind of outline of the "tree", with bits and pieces here and there that show similarities between groups that would otherwise (based on modern species) be regarded as completely distinct. You are setting the bar very high. That's fair, because scientists are supposed to be skeptical.
Given the circumstances, you might legitimately wonder: why would we be able to say anything at all about ancient biological history from fossils? Here's the rub: although it should be rare, although it is unlikely, from time to time something should turn up. If the tree exists at all -- i.e. if it really is a succession of branches -- it is being sampled, however sparsely. As you sample it, the branches should start converging. The gaps should get smaller as they get more numerous. Furthermore, you don't need every single twig from a tree to get a sense of it's broad shape, and occasionally you might get lucky and find a bit from very close to a major branch poin
Re:Yes, that Lenski (Score:3, Informative)
Another account of the story is at RationalWiki: http://rationalwiki.com/wiki/Lenski_affair [rationalwiki.com]
RationalWiki is a site that exists to poke fun at Conservapedia and the anti-science movement. (I particularly like its WIGO page, "What Is Going On At CP? [rationalwiki.com]".) Conservapedia forbids any mention of RationalWiki, going so far as to ban members who make oblique references to it. In fact, the part of Lenski's letter that was censored on Conservapedia as "Ed.: citation omitted due to spam filter" was, originally, a reference to RationalWiki; this is explained at "Censoring of Lensku's RW ref [rationalwiki.com]".